Exploring the link between initial and final diagnosis in a medical intelligent tutoring system

A constant topic in medical education is clinical reasoning: how do learners solve cases? Learner interactions with Intelligent Tutoring Systems yield fine-grained data that are useful in generating meaningful information and illuminating understanding about learner behaviors and outcomes. We examine and analyze the log files generated by BioWorld, an Intelligent Tutoring System for the medical domain. More specifically, to further our understanding of the nature of reasoning employed by learners while solving virtual patient cases in BioWorld, one important step is to examine the initial list of selected diagnostic hypotheses before any other learner action is taken in diagnosing a case. By exploring the link between initial selected hypotheses and final submitted hypothesis, a better understanding of the learners' reasoning might be achieved.